Molecular Docking Analysis on the Designed Benzimidazole Derivatives as

http://doi.org/10.17576/jsm-2023-5204-13

Molecular Docking Analysis on the Designed Benzimidazole Derivatives as EGFR Inhibitors: Comparison between EGFR Wild-Type ( EGFR

_WT

) and T790M Mutant

(Analisis Dok Molekul pada Terbitan Benzimidazol Direka sebagai Perencat EGFR: Perbandingan antara Jenis Liar EGFR (EGFR_WT) dan Mutan T790M)

NURUL AWANI SYAZZIRA JALIL¹ & SHAFIDA ABD HAMID^1,2,*

1Department of Chemistry, Kulliyyah of Science, International Islamic University Malaysia, 25200 Bandar Indera Mahkota, Kuantan, Pahang Darul Makmur, Malaysia

2SYNTOF, Kulliyyah of Science, International Islamic University Malaysia, 25200 Bandar Indera Mahkota, Kuantan, Pahang Darul Makmur, Malaysia

Received: 15 August 2022/Accepted: 26 February 2023

ABSTRACT

The non-small cell lung (NSCL) and colorectal cancers are frequently linked with the oncogenic activation of the epidermal growth factor receptor (EGFR), a member of the receptor tyrosine kinase (RTK) family. Current tyrosine kinase inhibitors (TKIs) are susceptible to drug resistance mutations and induce cytotoxicity effects on normal EGFRs.

The isosteric nature of benzimidazole with purine renders its great potential to imitate the binding mode of the purine-based ATP and prevents its contact with the EGFR active sites. Here, we report the molecular docking of 50 designed benzimidazole derivatives, as well as Gefitinib and ATP, to analyse and compare their binding modes at EGFR_wt and T790M active sites. The design of the ligands is based on our previous study, in which we proposed to evaluate keto- and amino-benzimidazoles, attached to a double bond linker and a phenyl group having electron donating and electron withdrawing groups attached at various positions. Docking simulations showed that keto-benzimidazoles dominated the top ten highest binding affinities in both EGFR-ligand complexes. The presence of sulfonyl substituents contributed to more stable complexes compared to others with binding energies of -8.1 (7c) and -7.8 (11c) kcal/mol in ^EGFR_wt, and -8.3 (7d) and -8.4 (1c) kcal/mol for T790M mutant. The substituent effects on the benzimidazole contributed not only to the hydrogen bonding and hydrophobic interaction, but also to the often-disregarded Van der Waals forces that are responsible for shape complementarity of the benzimidazoles with the EGFR binding pocket.

Keywords: Benzimidazole; EGFR; molecular docking; tyrosine kinase inhibitor; T790M

ABSTRAK

Sel paru-paru bukan kecil (NSCL) dan kanser kolorektum sering dikaitkan dengan pengaktifan onkogenik reseptor faktor pertumbuhan epidermis (EGFR), ahli keluarga reseptor tirosin kinase (RTK). Perencat tirosin kinase (TKI) masa kini terdedah kepada mutasi rintangan dadah dan mendorong kesan kesitotoksikan pada EGFR normal.

Sifat isosterik benzimidazol dengan purin menyebabkan ia berpotensi tinggi untuk meniru mod pengikatan ATP berasaskan purin dan menghalang sentuhannya dengan tapak aktif EGFR. Di sini, kami melaporkan dok molekul bagi 50 terbitan benzimidazol yang direka berserta Gefitinib dan ATP untuk menganalisis dan membandingkan mod pengikatan mereka di tapak aktif EGFR_wt dan T790M. Reka bentuk ligan adalah berdasarkan kajian terdahulu kami dan kami mencadangkan untuk menilai benzimidazol-keto dan amino- yang terikat pada penghubung ikatan berganda dan kumpulan fenil yang mempunyai kumpulan pendermaan elektron dan penarikan elektron yang dilampirkan pada pelbagai kedudukan. Simulasi dok mendedahkan bahawa benzimidazol-keto menguasai sepuluh pertalian pengikatan tertinggi dalam kedua-dua kompleks ligan EGFR. Kehadiran penukarganti sulfonil menyumbang kompleks yang lebih stabil berbanding yang lain; dengan tenaga pengikatan -8.1 (7c) dan -7.8 (11c) kcal/mol dalam EGFR_wt dan -8.3 (7d) dan -8.4 (1c) kcal/mol untuk mutan T790M. Kesan penukarganti pada benzimidazol menyumbang bukan sahaja kepada ikatan hidrogen dan interaksi hidrofobik, tetapi juga kepada daya Van der Waals yang sering diabaikan, yang bertanggungjawab untuk saling melengkapi bentuk benzimidazol dengan poket pengikat EGFR.

Kata kunci: Benzimidazol; dok molekul; EGFR; perencat tirosin kinase; T790M

INTRODUCTION

Cancer is caused by unregulated cell proliferation forming either a benign or malignant tumour. A malignant tumour can continuously spread through the bloodstream, infects other tissues and disrupts bodily functions, causing fatality (Mathur et al. 2015;

Nussbaumer et al. 2011; Rebucci & Michiels 2013).

The Global Cancer Statistics reported 19.3 million new cancer cases in 2020 and estimated a 47% rise by 2040 (Sung et al. 2021). The occurrence of non-small cell lung cancer (NSCL) and colorectal cancer is frequently linked with the oncogenic activation of EGFR. The metastasis can be curbed by targeting and inhibiting the oncogenic epidermal growth factor receptor (EGFR) residing on the cellular membrane (Troiano et al. 2016). EGFR is one of a family of four receptor tyrosine kinases (RTKs) in humans, the others being ErbB2/HER2, ErbB3/HER3, and ErbB4/HER4, which are responsible for initiating complex biological functions such as cell growth, motility, proliferation, and apoptosis by passing intercellular signals. Conformational changes caused by the binding of ligands such as growth factors and hormones to the extracellular domain of RTKs induce their dimerisation.

The dimer is only activated once the phosphate group from ATP is transferred to the tyrosine on the C-terminal of the RTK domain. The phosphorylated tyrosine then engages the neighbouring relay proteins, which would give out signals for specific cellular responses (Amelia et al. 2022).

Inhibitors of these receptors, involving the blocking of the EGFR binding site from interacting with agonists have been among the most successful examples of targeted cancer therapies. The action triggers the metastasis stage of malignant tumour cells by silencing the overexpressed or mutated EGFR. There are currently three TKIs generations clinically practiced, all of which bear an azaheterocyclic moiety. However, EGFR-target therapy drugs are continuously being developed to overcome the drugs’ limitations in terms of toxicity, non- selectivity, and resistance to mutations.

Benzimidazole is non-foreign in pharmacology as its derivatives are frequently associated with various biological activities. The scaffold is a structural isostere of indole and purine and thus its derivatives are expected to exhibit good/favourable affinity with various types of receptors. Benzimidazole derivatives exhibit anticancer activity via a few mechanisms, such as through topoisomerase I and II inhibition, DNA intercalation, PARP-poly inhibition, dihydrofolate reductase (DHFR)

and aromatase inhibition (Cevik et al. 2022; Karadavi et al. 2020; Mostafa, Gomaa & Elmorsy 2019). For example, the derivatives have been employed in two known cancer drugs, which are Veliparp and Nocodazole, although both drugs act via poly(ADP ribose) polymerase (PARP) inhibition and microtubule function disruption, respectively (Shrivastava et al. 2017). Benzimidazole interacts with Met793 of EGFR via a similar binding mode to quinazoline, owing to the nitrogen atoms on the nucleus acting as hydrogen bond acceptors (Abdullah, Ali & Hamid 2021). Issues such as cytotoxicity, non- selectivity, and unforeseen EGFR gene mutations further demand a more potent chemotherapeutic agent.

Given that benzimidazole is structurally similar to quinazoline, the scaffold for first- and second-generation drugs, the compound could be potentially served as an effective nucleus for future EGFR antagonists. However, there are still lots of studies that need to be carried out before the benzimidazole-based drug can be clinically approved for EGFR inhibition applications. Majority of the studies on structural insights of benzimidazole derivatives showed ambiguous findings on the type and position of the groups attached to the main scaffold, as well as the superiority of electron withdrawing and donating groups incorporation in improving the EGFR binding activity.

Overexpression of the EGFR wild-type (EGFR_wt) occurs more frequently than in the EGFR mutants (Thomas & Weihua 2019). The EGFR (Figure 1) consists of 28 exons and its mutations are grouped into three classes of nucleotide changes; i) short deletions of residues between Glu746 and Ser752 encoded by exon 19 (Class I), ii) substitution between exons 18 to 21 (Class II), and iii) duplications or insertions in exon 20 (Class III) [6]. These mutations cause the ligands to activate the tyrosine, exerting aberrant intercellular signals that trigger malignant tumour growth (Zandi et al. 2007).

The current tyrosine kinase inhibitors (TKIs) exhibited only initial clinical responses before the malignant tumour develops resistance towards the inhibition, and the efficacy of first and second generation TKIs, namely Gefitinib and Erlotinib, are often disrupted by the so- called T790M ‘gatekeeper’ mutation. The mutation at the EGFR tyrosine kinase domain has been described to be caused by the substitution of threonine with a bulky methionine, which results in steric interference for the TKIs to engage with the pocket. Further studies showed the mutation also increases drug resistance by increasing the ATP binding affinity and thus weakens the binding affinity of the drugs (Vyse & Huang 2019).

The third generation TKI such as Osimertinib addresses the T790M mutation using the flexibility of its structure to pass through the entrance and anchors onto the binding site (Lee at al. 2019). Despite being able to sensitize for the T790M EGFR, Osimertinib is inactive against novel mutation on EGFR L792 and C797S (Grabe et al. 2018; Jiang et al. 2018; Leonetti et al. 2019;

Song & Yang 2022). This leads to the development of the fourth generation TKIs such as EAI045 to inhibit the kinase activity of double mutant T790M/C797S.

EAI045 however must be used in combination with another anti-EGFR agent, namely Cetuximab (Thomas

& Weihua 2019). The resistance mechanism becomes more challenging when multiple mutations are involved.

STRUCTURE-ACTIVITY RELATIONSHIP BETWEEN BENZIMIDAZOLE AND EGFR

The biological activities exhibited by benzimidazoles prompted continuous studies on their structure-activity relationship to aid drug discovery. The activities are typically elicited when the substituents are grafted at 1,2,4 and/or 5 positions (Figure 3) (Bansal & Silakari 2012), although most studies concentrated on the attachment of various types of groups at C2, including alkyl chain, amide, carbonyl, benzene and heteroatomic ring. In many cases, these groups are also attached to substituted or unsubstituted phenyl group (Akhtar et al. 2017; Celik et al. 2019; Lelais et al. 2016; Srour et al. 2020).

Srour et al. (2020) designed benzimidazoles attached to thiazole and hydrazinylbenzene groups at C2 and found that the EGFR inhibitory potency decreases with a decrease in polarity of the molecules. In particular, derivatives containing p-nitrophenyl substituent showed significant EGFR activity. The importance of phenyl group linked to C2 of benzimidazole derivatives is also notable in many other studies, in which better EGFR inhibitory activity was exhibited by the phenyl group compared to the aliphatic chain (Akhtar et al. 2017; Celik et al. 2019). Furthermore, electron-rich substitution at the

para-position of a phenyl substituent was also found to increase the EGFR inhibitory activity (Akhtar et al.

2018; Labib et al. 2018). Several studies also relate the solubility of the compounds depends on the electronic property of the attached functional group. Furthermore, in many cases, the studies also keep the benzene side of benzimidazole unsubstituted while the few who did the opposite reported good EGFR inhibition. Therefore, insight into the relationship of these components and their binding activity on EGFR is imperative.

MATERIALS AND METHODS PROTEIN AND LIGAND PREPARATION

The crystal structures for EGFR_wt (PDB ID: 3VJO) and T790M mutant (PDB ID: 2JIT) were extracted from RSCB Protein Data Bank http://www.rcsb.org. The docking was carried out on monomeric form of the crystal structure. The AMP-PNP complexed with crystal structure was dismissed to leave a clean docking canvas.

The PDB files of the protein and ligand were prepared using AutoDock Tools version 1.5.7 (ADT; Scripps Research Institute, La Jolla, San Diego, USA). The three-dimensional protein target was stripped off all water molecules. Polar hydrogens were integrated into the proteins to allow establishment of hydrogen bonding during docking. Kollman and Gasteiger charges were also added to reconstruct the molecular electrostatic potential. Kollman charges were computed based on quantum mechanics whereas Gasteiger charges were generated based on electronegativity equilibration. Fifty benzimidazole analogues were designed and sketched using Chemsketch following previous published work (Abdullah, Ali & Hamid 2021) and constructed as in Figure 1. The SMILES notations were then converted into 3D structures and auto-optimized based on Universal Force Field (UFF) using Avogadro. The SDF file of ATP and Gefitinib were extracted from PubChem (https://

pubchem.ncbi.nlm.nih.gov) and prepared using Autodock Tools.

FIGURE 1. The structure of the designed benzimidazole derivatives

MOLECULAR DOCKING

AutoDock Vina version 1.2.3 software was used to perform the molecular docking simulations (The Scripps Research Institute, La Jolla, San Diego, USA). Discovery Studio Biovia 2021 (Dassault Systèmes, San Diego, California, USA) was employed to visualize and modify receptor and ligand structures. All 50 benzimidazole analogues, ATP and Gefitinib were docked at the predicted binding site. We have also superimposed our docking results (3VJO with gefitinib) with 4WKQ (crystal structural of EGFR of kinase domain with gefitinib). The protein superimpose returns an RMSD value of 0.473, which is acceptably low, deeming the docking method as trustable (López-Camacho et al. 2016). The residues making up the binding and active site for human EGFR (P00533), extracted based on the UniProtKB database were found to be Lys745, Asp837, and Asp855 (Yun et al. 2008). The protein grid box, which is the active site for docking was set up using AutoDock Tools to enclose the aforementioned residues. The grid box dimensions for T790M mutant at 0.375 Å are 36 × 44 × 48 and centered at -14.508 × 28.058 × 28.036 while for EGFR_wt at 0.375 Å are 40 × 60 × 34 and centered at 51.377 × 3.559 × -25.806. The docking simulations were then performed using AutoDock Vina, where the docking scores (in kcal/mol) were generated, and the binding poses ranked from the lowest to the highest according to the order of binding affinity. The energy range and exhaustiveness for the docking procedure were fixed at 4 and 32, respectively. The protein-ligand interactions were visualized with Biovia Discovery Studio Visualizer and the hydrogen and hydrophobic interactions within the protein-ligand complex were identified.

RESULTS AND DISCUSSION

The main approach in the development of EGFR inhibitors is still based on the currently used drugs, employing both bioactivity and bioactivity analyses.

Because of the diverse types of groups attached to different positions of the scaffold, structural insights of benzimidazole derivatives showed conflicting findings, especially on the superiority of electron withdrawing and donating groups in improving EGFR binding activity.

A closer look at the literature on the docking studies of benzimidazole derivatives with the EGFR protein showed that hydrogen bonds are mainly centralised at the middle part of the structures having an amino or/

and carbonyl group, and hydrophobic interaction at the phenyl group of the benzimidazole ring (Abdullah,

Ali & Hamid 2021). Thus, we proposed to evaluate benzimidazoles containing an amino or a carbonyl group attached to C2, attached to a double bond linker and a phenyl group. In this study, we designed 50 keto- and amino-benzimidazole derivatives having electron withdrawing groups; nitro, sulfonyl fluoride, sulfonate and electron donating group; methoxy, grafted onto the ortho-, meta- and para- position of the phenyl moiety and on C4, C5 and C7 of the benzimidazole nucleus (Table 1). The structures were optimised and docked against the wild-type EGFR (EGFR_wt; PDB ID: 3VJO) and EGFR mutant (T790M; PDB ID: 2JIT).

The EGFR-benzimidazole contact analysis shows the analogues interact with the EGFR_wt at the ATP and allosteric binding sites, while the binding pocket targeted within T790M mutant is only the ATP binding site (Figure 2). The binding affinity of the compounds with EGFR_wt and T790M mutant is shown in Table 2.

The benzimidazole derivatives exhibited better binding affinity with T790M mutant than EGFR_wt. Eight derivatives with the lowest binding free energy with EGFR_wt (-7.8 to -8.1 kcal/mol) occupied the allosteric binding site, suggesting the potential of these ligands as a Type II inhibitor. This observation is attributed mainly to the hydrogen bond interaction of Asp855 within the Asp-Phe-Gly (DFG) motif of the activation loop with the oxygen atoms of carbonyl linker, sulfonyl fluoride, sulfonate and nitro moieties. The DFG motif is responsible for directing Mg²⁺ into the catalytic site to interact with the phosphate groups in ATP (Amelia et al. 2022). The oxygen atom of sulfonyl fluoride and methoxy group also act as Michael acceptors and approach the Cys797 residue in the hinge region of T790M mutant for hydrogen bond formation (7a, 7c, and 11c). Inhibitors such as HKI-272 and AST1306 attribute their irreversibility to the covalent bond formations with Cys797, hence similar binding mode could potentially render the benzimidazole analogues irreversible and able to silence the tumour cells permanently (Xie et al. 2011; Yun et al. 2008).

The N1 of the imidazole ring becomes a hydrogen donor for Met793, Leu777 and principally Thr854, while the N3 of the ring frequently interacts with Lys745 and Leu703 through hydrogen bonding. On the other hand, benzimidazole nucleus and substituted phenyl hydrophobically interact with Val726, Phe723, Ala743, Leu844, Met790 and Leu718 of both receptors (Table 2). These interactions indicate the frequent contact of benzimidazole analogues with the important residues for ATP binding, namely Lys745, Leu718, Val726, Ala743 and Leu844. Blocking these residues from ATP could

TABLE 1. The designed benzimidazole derivatives

Ligand X R₁/R₂

position R₁ R₂ Ligand X R₁/R₂

position R₁ R₂ 1a

NH

(para)C15

-OCH₃ H 7a -OCH₃ H

1b -NO₂ H 7b C15 -NO₂ H

1c -SO₂F H 7c

CO

(para) -SO₂F H

1d -SO₃H H 7d -SO₃H H

2a

(meta)C14

-OCH₃ H 8a -OCH₃ H

2b -NO₂ H 8b C14

(meta) -NO₂ H

2c -SO₂F H 8c -SO₂F H

2d -SO₃H H 8d -SO₃H H

3a

(ortho)C13

-OCH₃ H 9a

(ortho)C13

-OCH₃ H

3b -NO₂ H 9b -NO₂ H

3c -SO₂F H 9c -SO₂F H

3d -SO₃H H 9d -SO₃H H

4a

C5

H -OCH₃ 10a

C5

H -OCH₃

4b H -NO₂ 10b H -NO₂

4c H -SO₂F 10c H -SO₂F

4d H -SO₃H 10d H -SO₃H

5a

C4

H -OCH₃ 11a H -OCH₃

5b H -NO₂ 11b C4 H -NO₂

5c H -SO₂F 11c H -SO₂F

5d H -SO₃H 11d H -SO₃H

6a

C7

H -OCH₃ 12a H -OCH₃

6b H -NO₂ 12b C7 H -NO₂

6c H -SO₂F 12c H -SO₂F

6d H -SO₃H 12d H -SO₃H

13 - - - 14 - - -

1

2 4 3 5 6

7 8

9

10 11d

12 13

14 15

16 17

potentially suppress signal transduction pathways and cellular responses. The nitrogen atoms of benzimidazoles with carbonyl linker (7a, 7b, 7c, 7d, 8a, 8c and 8d) consistently form hydrogen bonds with Thr854 and Lys745 of EGFR_wt. The benzimidazole nucleus also forms π-anion interaction with Asp855 while the distal phenyl form π-interaction with Val726 and Ala743.

When the substituents are grafted onto the distal phenyl, the differences in the bond distance

gradually declined across the ligand body towards the benzimidazole nucleus (Figure 3). This finding may be due to localised interaction, that is when there is more than one aromatic ring in a system, the substituent effect is prominent on the nearest vertex and not wholly on the other rings (Wheeler 2012). The results also showed that benzimidazoles with carbonyl linkers primarily dominate the top ten highest binding affinities in both EGFR-ligand complexes (90% for EGFR_wt and 80% for T790M mutant).

FIGURE 2. ATP (green) and allosteric (blue) binding sites for benzimidazole analogues in EGFR_wt (left) and T790M mutant (right)

TABLE 2. EGFR_wt-ligand and T790M-ligand contact analysis from highest to lowest binding affinity (kcal/mol)

EGFR_wt T790M

Ligand Binding affinity (kcal/mol)

of H No bonds

H bond residues

Hydrophobic interacting

residues Ligand Binding affinity (kcal/mol)

of H No bonds

H bond

residues Hydrophobic interacting residues

7c -8.1 4

Lys745, Asp855 Thr854, Met793

Ala743

Val726 1c -8.4 4

Asp837 Asp855 Thr854 Asn842

Lys875 Leu858,Val726

11c -8.0 3 Lys745,

Asp855

Leu718 Leu844, Val726,

Ala743, Lys745 7d -8.3 1 Asn842 Ala743, Met790

Leu844, Phe723

12c -8.0 2 Asp855 Lys745

Val726, Leu718,

Ala743, Leu844 8c -8.3 4

Asp837 Asp855 Asn842 Thr854

Lys875, Leu858 Val726, Lys745

12d -8.0 3 Lys745,

Asp855

Leu718, Leu844 Val726, Ala743

Lys745 10c -8.2 2 Lys745 Lys875, Ala722

Phe723, Val726

11d -7.9 4 Lys745, Asp855 Glu762

Leu844, Leu718 Val726, Lys745

Ala743 11c -8.2 2 Asn842

Cys797 Phe723, Leu844 Ala743

5d -7.8 4 Asp855,

Lys745

Leu718, Leu844 Ala743, Val726

Lys745 4c -8.2 0 - Ala743, Leu844

Phe723

6d -7.8 4 Asp855,

Lys745 Leu718, Leu844

Val726, Ala743 7c -8.2 1 Cys797 Phe723, Lys745

Val726, Ala743 Leu844

8c -7.8 4 Met793,

Lys745 Thr854 Ala743, Val726 10d -8.1 3 Arg836

Lys860 Leu858, Ile759 Ala755

9c -7.8 3 Tyr1016

Ile1018,

Leu777 Leu778, Ile1018 12b -8.1 3 Thr854

Asn842 Asp855

Ala743, Leu844 Phe723

9d -7.8 5 Arg776,

Leu703 Arg705, Leu778

Arg776 12d -8.1 2 Gln791

Asn842 Leu844, Ala743 Val726, Lys745

10c -7.7 4 Leu703,

Ile1018

Arg776 - 5d -8.1 3 Arg841

Asp855 Val726, Phe723 Leu858, Lys875

11b -7.7 3 Asp855,

Lys745

Leu718, Val726 Ala743, Lys745

Leu788 6b -8.1 0 - Pro877, Lys875

Leu858, Lys745 Val726

12b -7.7 2 Thr854,

Asp855 Lys745, Val726

Ala743, Leu718 7b -8.1 2 Asn842

Lys745 Ala743, Leu844 Met790, Phe723

5c -7.7 3 Asp855,

Lys745

Leu718, Leu844 Ala743, Val726

Lys745 8d -8.1 0 - Phe723,Val726

Leu718

7d -7.7 4

Lys745, Asp855 Thr854, Met793

Val726, Ala743 9b -8.1 2 Thr854

Asn842

Phe723, Lys745 Val726, Ala743

Leu844

8d -7.7 3 Lys745,

Thr854

Met793 Ala743, Val726 10b -8.0 1 Lys745 Phe723, Val726

Leu718, Ala743 Leu844

9b -7.7 4 Arg776,

Leu703 Ile1018 11d -8.0 2 Asn842

Asp855 Leu844 ,Ala743 Phe723

2c -7.6 5

Val769, Arg776 Ala767, Ile1018

Ile1018, Ala702

Gln701 12c -8.0 1 Asn842 Leu844, Phe723

8b -7.6 2 Asp855,

Met793 Lys745, Val726

Ala743, Thr790 1d -8.0 1 Asp855 Ala743, Leu844

Val726, Phe723

10b -7.5 2 Met793,

Thr654 Leu844, Ala743

Val726 2c -8.0 0 - Leu844, Ala743

Val726, Phe723

10d -7.5 2 Thr854,

Met793 Val726, Leu844

Ala743 2d -8.0 0 - Ala743, Leu844

Met790, Val726 Phe723

1c -7.5 4

Glu762, Lys745 Met793,

Thr854

Ala743, Val726 3c -8.0 0 - Leu844, Val726

Ala743, Phe723 Lys745

2d -7.5 3 Arg776,

Leu703

Leu777 Arg705, Leu778 6c -8.0 1 Asp855 Lys745, Leu858

Pro877, Lys875 Val726

3b -7.5 4 Ile1018,

Arg776

Leu703 Ala702, Leu778 8b -8.0 2 Asp855

Thr854 Phe723, Val726 Leu844, Ala743

3c -7.5 3 Arg776,

Leu777

Leu703 Leu778 9c -8.0 2 Thr854 Phe723, Lys745

Val726, Ala743 Leu844

3d -7.5 4 Arg776,

Leu777

Leu703 Arg705 9d -8.0 0 - Phe723, Lys745

Val726, Ala743 Leu844, Leu718

7b -7.5 4

Lys745, Asp855 Thr854, Met793

Ala743, Val726 11b -7.9 1 Asn842

Asp855 Leu844, Ala743 Phe723

5b -7.4 1 Asp855 Leu718, Val726

Val743, Ala743 1b -7.9 0 - Ala743, Leu844

Met790, Val726 Phe723

6b -7.4 1 Lys745 Ala743, Leu718

Leu844, Lys745 3d -7.9 2 Asp855

Thr854

Val726, Phe723 Lys745, Leu844 Ala743, Leu718

1d -7.3 3 Lys745,

Thr854

Glu762 Val726, Ala743 5c -7.9 1 Asp855 Leu844, Ala743

Met790, Phe723

4b -7.3 2 Thr790,

Met793

Val726, Lys745 Ala743, Cys775

Leu844 3b -7.8 3 Asp855

Thr854 Asn842

Phe723, Leu844 Ala743

6c -7.3 2 Leu777,

Leu703 Leu778 4d -7.8 0 - Phe723, Leu844

Ala743

ATP -7.3 13

Thr790, Thr854 Asp855,

Lys745 Arg841, Asn842 Met793

Leu844 6d -7.8 1 Asn842 Leu844, Ala743

Phe723

aGEF -7.3 1 Lys745 Ala743, Val726

Leu844, Cys797 7a -7.8 1 Cys797 Lys745, Val726

Leu844, Leu718 Cys797

2b -7.2 3 Met793,

Thr854 Leu718, Val726

Ala743 10a -7.7 1 Lys745 Val726, Phe723

Leu718, Leu844

4c -7.2 3 Arg705,

Leu703 Ala767

Ala702, Leu703

Arg705 14 -7.7 0 - Lys745, Val726

Leu718, Leu792 Leu844

4d -7.2 5

Ile1018 Tyr1016 Arg776, Ala767 Leu703

Gln701, Ala702 11a -7.7 1 Asn842 Ala743, Leu844

Met790, Leu792 Phe723

9a -7.2 3 Arg776 Ile1018 12a -7.7 1 Asn842 Ala743, Leu844 Met790, Leu792 Leu718, Phe723

10a -7.1 1 Met793 Val726, Ala743

Leu844, Leu788

Met766 3a -7.7 0 - Leu844, Ala743

Met790, Val726 Phe723

12a -7.1 2 Arg776,

Leu777

Leu1017 Leu778, Ile1018

Arg705, Leu703 13 -7.6 0 - Leu844, Ala743

Met790, Val726 Phe723

11a -7.0 2 Arg776

Tyr1016 Ile1018, Leu703

Arg705 2b -7.6 1 Lys745 Phe723, Val726

Ala743, Leu844

13 -7.0 1 Thr854 Leu788, Thr790

Ala743, Leu718

Val726 5b -7.6 0 - Lys745, Val726

Leu792, Leu718 Leu844

14 -7.0 0 - Leu788, Ala743

Thr790, Lys745 8a -7.6 0 -

Phe723, Lys745 Val726, Leu844 Ala743, Leu718

Cys797

5a -7.0 4

Ile1018, Leu778 Leu777, Leu703

Leu1017

Leu778 ^bGEF -7.6 1 Lys745 Val726, Lys745

Phe723, Leu858

7a -7.0 3 Lys745,

Asp855 Thr854

Met793, Leu844

Val726, Ala743 1a -7.5 1 Asp855 Lys875, Pro877

Leu858, Lys745 Val726

8a -7.0 4 Lys745,

Asp855 Thr854

Val726, Ala743

Met793, Leu844 4a -7.5 2 Lys745,

Thr854 Leu792, Leu844 Phe723

1b -6.8 3 Lys745,

Thr854

Glu762 Val726, Ala743 4b -7.5 1 Lys745 Leu844, Ala743

Phe723, Val726 Lys745

6a -6.8 1 Thr790 Lys745, Val726

Ala743, Leu718

Leu844 9a -7.5 0 - Phe723, Lys745

Val726, Leu844 Ala743, Leu718

1a -6.7 1 Thr854 Lys745, Val726

Thr790, Ala743

Leu718 6a -7.4 1 Asp855 Leu718, Leu792

Leu844, Met790 Ala743, Phe723

2a -6.5 4 Thr790,

Met793 Pro794

Ala743, Leu844 Met766

Val726, Leu718 5a -7.3 1 Asp855 Phe723

Met790, Leu844 Leu792, Leu718

3a -6.5 2 Leu777,

Leu703 Ala702, Arg705

Leu778 2a -7.2 0 - Phe723, Val726

Met790, Leu844 Ala743

4a -6.5 0 - Leu718, Leu844

Ala743, Val726 ATP -7.2 2 Asn842

Gln791 Leu844, Met790 Ala743

aGEF = Gefitinib

The binding energies of unsubstituted amino- (13) and keto-benzimidazole (14) are -7.0 kcal/mol for EGFR_wt and -7.6 and -7.7 kcal/mol, respectively, for T790M mutant. The presence of sulfonyl substituents significantly gave good protein-ligand complex binding affinities as shown by the lower binding free energies (more stable complexes) compared to others; -8.1 (7c) and -7.8 (11c) kcal/mol in EGFR_wt complex, and -8.3 (7d) and -8.4 (1c) kcal/mole for T790M mutant. These observations may be attributed to the strong electron withdrawing effect of both moieties. Sulfonyl fluoride is expected to reduce the electron density of its phenyl more than sulfonate as fluorine is more electronegative than oxygen, and thus attenuates the π-alkyl interactions between the aromatic ring of benzimidazole, and Val726 and Ala743 of the protein (Liégeois et al. 2014; Neel et al. 2017). In this study, however, the fluorine effect is not prominent enough as the binding free energies for both sulfonyl fluoride and sulfonate are quite similar. Nitro substituent, being slightly weaker electron withdrawing than sulfonyl fluoride and sulfonate showed lower binding affinity when docked against both EGFR receptors.

Methoxy is the only substituent exerting electron donating effect among the substituents and consistently gave the highest binding free energies in the range of -6.5 to -7.2 kcal/mol in EGFR_wt and -7.2 to -7.7 kcal/mol for T790M mutant. Hence, we postulated that electron withdrawing groups contribute to the protein-ligand interaction better than electron donating groups. In addition, the binding affinity also correlates with the strength of the electron withdrawing group (-SO₂F = -SO₃H > -NO₂ >

H > -OCH₃). Interestingly, no correlation was observed between the binding affinity of EGFR-ligand complex

with the C4, C5 and C7 position of benzimidazole and the para, ortho- and meta- position of the substituted phenyl.

COMPARISON OF DOCKING RESULT FROM BENZIMIDAZOLE, ATP AND GEFITINIB

Despite having 13 hydrogen bonds with distances ranging from 2.07 to 3.05 Å, the ATP-EGFR_wt complex gave a binding energy of only -7.3 kcal/mol, while benzimidazole analogues docked against the same protein showed a maximum of -8.1 kcal/mol binding free energy with only four hydrogen bonds at 1.83 to 2.48 Å distance range. Similarly, the significantly different binding energy is also emulated in the ATP-T790M mutant complex when compared with the benzimidazole derivatives against the T790M mutant. This deviation signifies that binding affinity does not necessarily correlate with the number of hydrogen bonds. Instead, the lower binding free energy of the benzimidazole analogues could be attributed to the surface shape complementarity that is governed by the Van der Waals interactions of the docked ligands (Tsai et al. 2001). Although hydrogen bond interactions played a major role in stabilising the ligands at the binding interface, the complementary shape between proteins and their ligands is also critical for predicting protein-ligand binding affinities as it governs the attraction and repulsion between non-bonding atoms (Hsu, Chen & Yang 2008).

Unlike ATP, the benzimidazole derivatives mostly adopt a conformation that complements the shape of the binding pocket (Figure 4). As torsional degrees of freedom were not assigned to the protein and ligands during the docking procedure, it is assumed that the interactions in this study follow a ‘lock-and-key’

FIGURE 3. Interactions of (a) EGFR_wt-7c (left) and (b) EGFR_wt-7d (right)

1 2 4 3

5 6

7 8

9 10 11

12 13 14 16 15 17

17

16 15

14 13 12 11 10 2

1 9 3 4 5 6

7 8

model instead of the induced fit model. Hence, the complementary shape of the benzimidazoles would be a better ‘key’ for the binding pocket compared to ATP.

On the whole, benzimidazole analogues displayed lower binding free energy than Gefitinib. The binding affinity difference between Gefitinib and the

benzimidazole analogues scales at 0.5~0.8 kcal/mol in both EGFR_wt and T790M mutant. This observation could also be associated to the shape of the ligands within the receptors, as shown in Figure 5. Benzimidazoles were shown to possess a more complementary shape for the binding pocket of EGFR_wt and T790M mutant when compared to Gefitinib.

(a) (b)

FIGURE 4. Comparison of adopted conformation by benzimidazole analogue (green) and ATP (red) after being docked against (a) EGFR_wt and (b) T790M mutant

FIGURE 5. Comparison of adopted conformation by benzimidazole analogue (red) and Gefitinib (green) after being docked against (a) EGFR_wt and (b) T790M mutant

(a) (b)

CONCLUSIONS

Molecular docking analysis was used to evaluate the binding activity of 50 designed benzimidazole analogues toward EGFR_wt and T790M mutant active

sites. The analysis showed strong binding affinities of the derivatives against the target receptors, ranging from -4.77 to -8.30 kcal mol^-1. Our docking study also showed keto-benzimidazoles exhibiting the top ten

highest binding affinities against both EGFR active sites, ranging from -7.8 to -8.1 kcal/mol (for EGFR_wt) and -8.1 to -8.4 (in T790M). The top ten compounds also showed better binding affinity compared to Gefitinib and ATP in both active sites. The N1 of the imidazole ring acted as a hydrogen donor for Thr854, Met793 and Leu777, while Lys745 and Leu703 interacted with N3 via hydrogen bonds. The benzimidazole scaffold and substituted phenyl moiety interacted hydrophobically with Val726, Phe723, Ala743, Leu844, Met790 and Leu718 in both receptors. The strong electron withdrawing sulfonyl substituent contributed significantly to the protein-ligand complex binding affinity. However, the complementarity shape of the benzimidazoles with the binding pocket of EGFR_wt and T790M mutant also played a pivotal role in regulating the binding free energies. No correlation was seen between the binding free energy with the C4, C5 and C7 positions of benzimidazole and the para-, ortho- and meta- positions of the substituted phenyl. Intermolecular steric clashes were noticed in some complexes, namely 2a, 3b, 4d, 5c, 5d, 6b, 6d, 8b, 9d, 10c, 10d, and 12c for EGFR_wt-benzimidazole, and 5a, 5b, 6a, 6b, and 9d for T790M-benzimidazole. As the flexibility of both protein and ligand were not taken into account, their rigidity causes the ligands’ conformation to not fit perfectly into the binding pocket. Employing molecular dynamics could prevent these clashes and provide better binding poses as the software allows some torsional degree of freedom.

Although studies on the design of benzimidazoles as EGFR inhibitors have been presented by many authors, the matter is still insufficiently explored. This study could assist chemists in the design and synthesis of this class of benzimidazoles toward finding the next generation of EGFR inhibitors. The compounds can be further validated by in vitro and in vivo assays.

ACKNOWLEDGEMENTS

We are grateful to the Ministry of Higher Education (MOHE), Malaysia for the Fundamental Research Grant Scheme (FRGS/1/2018/STG01/UIAM/02/1). We would also like to thank Nurasyikin Hamzah for the useful discussion.

REFERENCES

Abdullah, M.N., Ali, Y. & Hamid, S.A. 2021. Insights into the structure and drug design of benzimidazole derivatives targeting the epidermal growth factor receptor (EGFR).

Chemical Biology & Drug Design 100(6): 921-934.

Akhtar, M.J., Khan, A.A., Ali, Z., Dewangan, R.P., Rafi, M., Hassan, M.Q., Akhtar, M.S., Siddiqui, A.A., Partap, S., Pasha, S. & Yar, M.S. 2018. Synthesis of stable benzimidazole derivatives bearing pyrazole as anticancer and EGFR receptor inhibitors. Bioorganic Chemistry 78:

158-169.

Akhtar, M.J., Siddiqui, A.A., Khan, A.A., Ali, Z., Dewangan, R.P., Pasha, S. & Yar, M.S. 2017. Design, synthesis, docking and QSAR study of substituted benzimidazole linked oxadiazole as cytotoxic agents, EGFR and erbB2 receptor inhibitors. European Journal of Medicinal Chemistry 126:

853-869.

Amelia, T., Kartasasmita, R.E., Ohwada, T. & Tjahjono, D.H. 2022. Structural insight and development of EGFR tyrosine kinase inhibitors. Molecules 27(3): 819.

Bansal, Y. & Silakari, O. 2012. The therapeutic journey of benzimidazoles: A review. Bioorganic & Medicinal Chemistry 20(21): 6208-6236.

Celik, I., Ayhan-Kılcıgil, G., Guven, B., Kara, Z., Gurkan- Alp, A.S., Karayel, A. & Onay- Besikci, A. 2019. Design, synthesis and docking studies of benzimidazole derivatives as potential EGFR inhibitors. European Journal of Medicinal Chemistry 173: 240-249.

Çevik, U.A., Celik, I., Mella, J., Mellado, M., Özkay, Y. &

Kaplancikli, A.Z. 2022. Design, synthesis, and molecular modeling studies of a novel benzimidazole as an aromatase inhibitor. ACS Omega 7(18): 16152-16163.

Grabe, T., Lategahn, J. & Rauh, D. 2018. C797S resistance:

The undruggable EGFR mutation in non-small cell lung cancer? ACS Medicinal Chemistry Letters 9(8): 779- 782.

Hsu, K-C., Chen, Y-F. & Yang, J-M. 2008. Binding affinity analysis of protein-ligand complexes. 2008 2nd International Conference on Bioinformatics and Biomedical Engineering.

pp. 167-171.

Jiang, T., Su, C., Ren, S., Cappuzzo, F., Rocco, G., Palmer, J.D., van Zandwijk, N., Blackhall, F., Le, X., Pennell, N.A.

& Zhou, C. 2018. A consensus on the role of Osimertinib in non-small cell lung cancer from the AME Lung Cancer Collaborative Group. Journal of Thoracic Disease 10(7):

3909-3921.

Karadayi, F.Z., Yaman, M., Kisla, M.M., Keskus, A.G., Konu, O. & Ates-Alagoz, Z. 2020. Design, synthesis and anticancer/antiestrogenic activities of novel indole- benzimidazoles. Bioorganic Chemistry 100: 103929.

Lee, Y., Kim, T.M., Kim, D.W., Kim, S., Kim, M., Keam, B., Ku, J.L. & Heo, D.S. 2019. Preclinical modelling of Osimertinib for NSCLC with EGFR Exon 20 insertion mutations. Journal of Thoracic Disease 14(9): 1556-1566.

Lelais, G., Epple, R., Marsilje, T.H., Long, Y.O., McNeill, M., Chen, B., Lu, W., Anumolu, J., Badiger, S., Bursulaya, B., DiDonato, M., Fong, R., Juarez, J., Li, J., Manuia, M., Mason, D.E., Gordon, P., Groessl, T., Johnson, K., Jia, Y., Kasibhatla, S., Li, C., Isbell, J., Spraggon, G., Bender, S. &

Michellys, P-Y. 2016. Discovery of (R,E)-N-(7-Chloro-1-(1- [4-(dimethylamino)but-2-enoyl]azepan-3-yl)-1H-benzo[d]

imidazol-2-yl)-2-methylisonicotinamide (EGF816), a novel, potent, and WT sparing covalent inhibitor of oncogenic (L858R, ex19del) and resistant (T790M) EGFR mutants for the treatment of EGFR mutant non-small-cell lung cancers.

Journal of Medicinal Chemistry 59(14): 6671-6689.

Labib, M.B., Philoppes, J.N., Lamie, P.F. & Ahmed, E.R.

2018. Azole-hydrazone derivatives: Design, synthesis, in vitro biological evaluation, dual EGFR/HER2 inhibitory activity, cell cycle analysis and molecular docking study as anticancer agents. Bioorganic Chemistry 76: 67-80.

Leonetti, A., Sharma, S., Minari, R., Perego, P., Giovannetti, E.

& Tiseo, M. 2019. Resistance mechanisms to Osimertinib in EGFR-mutated non-small cell lung cancer. British Journal of Cancer 121: 725-737.

Liégeois, J-F., Lespagnard, M., Salas, E.M., Mangin, F., Scuvée- Moreau, J. & Dilly, S. 2014. Enhancing a CH−π interaction to increase the affinity for 5-HT1A receptors. ACS Medicinal Chemistry Letters 5(4): 358-362.

López-Camacho, E., García-Godoy, M.J., García-Nieto, J., Nebro, A.J. & Aldana-Montes, J.F. 2016. A new multi- objective approach for molecular docking based on RMSD and binding energy. In Algorithms for Computational Biology. Lecture Notes in Computer Science, vol 9702, edited by Botón-Fernández, M., Martín-Vide, C., Santander- Jiménez, S. & Vega-Rodríguez, M. Springer, Cham.

Mathur, G., Nain, S. & Sharma, P.K. 2015. Cancer: An overview.

Academic Journal of Cancer Research 8(1): 1-9.

Mostafa, A.S., Gomaa, R.M. & Elmorsy, M.A. 2019. Design and synthesis of 2-phenyl benzimidazole derivatives as VEGFR-2 inhibitors with anti-breast cancer activity.

Chemical Biology Drug Design 93(4): 454-463.

Neel, A.J., Hilton, M.J., Sigman, M.S & Toste, F.D. 2017 Exploiting non-covalent π interactions for catalyst design.

Nature 543(7647): 637-646.

Nussbaumer, S., Bonnabry, P., Veuthey, J.L. & Fleury- Souverain, S. 2011. Analysis of anticancer drugs: A review.

Talanta 85(5): 2265-2289.

Rebucci, M. & Michiels, C. 2013. Molecular aspects of cancer cell resistance to chemotherapy. Biochemical Pharmacology 85(9): 1219-1226.

Shrivastava, N., Naim, M.J., Alam, M.J., Nawaz, F., Ahmed, S. & Alam, O. 2017. Benzimidazole scaffold as anticancer agent: Synthetic approaches and structure-activity relationship. Archiv. Der. Pharmazie. 350: e201700040.

Song, C. & Yang, X. 2022. Osimertinib-centered therapy against uncommon epidermal growth factor receptor-mutated non- small-cell lung cancer-A mini review. Frontiers in Oncology 12: 834585.

Srour, A.M., Ahmed, N.S., Abd El-Karim, S.S., Anwar, M.M.

& El-Hallouty, S.M. 2020. Design, synthesis, biological evaluation, QSAR analysis and molecular modelling of new thiazol-benzimidazoles as EGFR inhibitors.

Bioorganic Medicinal Chemistry 28(18): 115657.

Sung, H., Ferlay, J., Siegel, R.L., Laversanne, M., Soerjomataram, I., Jemal, A. & Bray, F. 2021. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: Cancer Journal for Clinicians 71(3): 209-249.

The UniProt Consortium. 2019. UniProt: A worldwide hub of protein knowledge. Nucleic Acids Research 47:

D506-D515.

Thomas, R. & Weihua, Z. 2019. Rethink of EGFR in cancer with its kinase independent function on board. Frontiers in Oncology 9: 800.

Troiani, T., Napolitano, S., Corte, C.M.D., Martini, G., Martinelli, E., Morgillo, F. & Ciardiello, F. 2016.

Therapeutic value of EGFR inhibition in CRC and NSCLC: 15 years of clinical evidence. ESMO Open 1(5): e000088.

Tsai, C-J., Norel, R., Wolfson, H.J., Maizel, J.V. & Nussinov, R.

2001. Protein-ligand interactions: Energetic contributions and shape complementarity. Encyclopedia of Life Sciences. pp. 1-8.

Vyse, S. & Huang, P.H. 2019. Targeting EGFR exon 20 insertion mutations in non-small cell lung cancer. Signal Transduction and Targeted Therapy 4: 5.

Wheeler, S.E. 2012. Understanding substituent effects in noncovalent interactions involving aromatic rings.

Accounts of Chemical Research 46(4): 1029-1038.

Xie, H., Lin, L., Tong, L., Jiang, Y., Zheng, M., Chen, Z., Jiang, X., Zhang, X., Ren, X., Qu, W., Yang, Y., Wan, H., Chen, Y., Zuo, J., Jiang, H., Geng, M. & Ding, J. 2011 AST1306, A novel irreversible inhibitor of the epidermal growth factor receptor 1 and 2, exhibits antitumor activity both in vitro and in vivo. PLoS ONE 6(7): e21487.

Yun, C-H., Mengwasser, K.E., Toms, A.V., Woo, M.S., Greulich, H., Wong, K-K., Meyerson, M. & Eck, M.J. 2008. The T790M mutation in EGFR kinase causes drug resistance by increasing the affinity for ATP. PNAS 105(6): 2070-2075.

Zandi, R., Larsen, A.B., Andersen, P., Stockhausen, M-T. &

Poulsen, H.S. 2007. Mechanisms for oncogenic activation of the epidermal growth factor receptor. Cell Signal 19(10):

2013-2023.

*Corresponding author; email: shafida@iium.edu.my